In laboratory research, the experimenter's role is to behave in a standardized manner, often following a script, in order to collect data without influencing the responses of subjects. However, in field studies of vulnerable or victimized populations, experimenters confront tasks and dilemmas at odds with this concept of their role. Subjects may cry; experimenters may experience strong feelings when witnessing adversity; normative experimental behavior may feel inappropriate. This paper reports on issues identified in a graduate seminar for nonclinically trained health psychology researchers, and describes the program of training provided. Paradoxically, adhering to a more detached, noninfluential style of interaction requires adapting a flexible, rather than a rote, behavioral style. Relevant skills include explicating values, developing relational and communications skills, and training in post-traumatic stress syndromes. By thus redefining the experimenter's role, ethical and practical considerations introduced in field studies of distressed populations can be balanced with laboratory values.